Error Impact of Regression Models on Forest Road Spacing

Statistical time prediction models are common in production estimation of different forest machineries. These models are usually developed using multiple regression method. Application of regression method causes considerable errors. In this paper, firstly the confidence interval curves of the parameters used in the yarding and installation time predicting models of tower yarder in Austria are presented.

The second part of this study deals with the effect of the error of regression models on optimal road spacing of tower yarder based on the minimization of total costs of roading, yarding and installation. Finally to choose the best optimal spacing under uncertainty, we used the multiple criteria decision making process considering the criteria like minimum total cost (Euro/m³) as economical criteria, soil erosion, soil compaction in the skid trail, area of constructed roads (m²/ha), reforestation costs and facilitate of silvicutural treatments to the stands.